5.1 SUMMARY OF THE PRESENT WORK
This work investigated the integrity and performance of a network as multiple heterogeneous systems, including robots and control software, were added to it. To perform this investigation, a virtual datacenter was built to support the creation and operation of software systems on virtual machines. The datacenter provided the computational resources necessary for the operation of any cloud-based software systems utilized by the robots. Furthermore, the datacenter provided the ability to build virtual machines as necessary to be the foundation of the software systems that were to be run.
The Robot Operating System was used as the backbone of the network communication that occurred from robot-to-cloud and inside the cloud. Through ROS, data streams were easily initialized, transmitted, and received throughout the network, passing along information to those agents that requested it. ROS nodes implemented on the virtual machines within the cloud provided cloud services to the robots, such as control systems to follow a black line or utilize the pose of one robot to control the velocity of another.
Varying amounts of data were passed from the robot to cloud and back, creating the load that the network performance and integrity was measured against. For each type of robot, baseline performance data was gathered, and then performance data for an increasing number of robots and an increasing amount of bandwidth usage. This data was processed to determine the latency of robot-to-cloud communications, as well as dropped data packets.
A cloud-in-the-loop control system was developed to determine the viability of implementing the controller of a time-sensitive system within the cloud. Though the functionality of the control system was preserved, the latency between the controller and the robot made the system unstable and unviable. Finally, a tilt control system was implemented to read from the pose of one robot
and interpret the data as velocity commands for another. The state of each robot was observed to determine the latency between maneuvering the control robot and receiving the velocity commands on the other.
The results confirm that it is possible to build a network that provides cloud services to heterogeneous client devices while maintaining integrity under some amount of load, though the integrity diminishes significantly as the wireless bandwidth capacity is saturated. The network supported many robots under low bandwidth applications, but few robots under high bandwidth applications.
5.2 FUTURE SCOPE OF WORK
Minimizing latency due to interference for research purposes would be useful for conducting very controlled experiments on robot performance in an isolated setting. To this end, two improvements could be made to the study: first, the use of a 5GHz wireless access point would result in much less interference, as the campus access points causing much of the interference are on the 2.4GHz band, as was the wireless access point used for this study. Second, the construction of a Faraday cage would provide the means to examine the performance of these devices in an environment completely isolated from outside interferences.
Additional networked systems could be integrated into the environment to provide supplementary sensor feedback, such as mounted 2D and 3D cameras. These cameras would provide additional network load, though the load could be through wired connections instead of wireless, increasing the capacity of the connection.
With IPv6 fast approaching implementation out of necessity due to a growing number of networked devices, the existing network could be switched to the IPv6 protocol to examine the challenges the protocol present and determine solutions to meet those challenges.
The performance of networked devices should be evaluated based on the ratio of local processing to remote processing, and the advantages and disadvantages of such a design evaluated. For those devices unable to be programmed, an intermediary device, such as a Field Programmable Gate Array (FPGA) could be added to handle some local processing to increase stability of the system.
REFERENCES
An, Baik Song, Manhee Lee, Ki Hwan Yum, and Eun Jung Kim. "Efficient data packet compression for cache coherent multiprocessor systems." Data Compression Conference, 2012. Snowbird, UT:
IEEE, 2012. 129-138.
Araos, Diego. "ardrone-wpa2." GitHub. March 29, 2013.
Armbrust, M, et al. "A view of cloud computing." Communications of the ACM 53, no. 4 (2010): 50-58.
Cao, Y., A. Fukunaga, and A Kahng. "Cooperative mobile robotics: antecedents and directions." Auton.
Robots, 1997: 1-23.
Deka, Ganesh Chandra. "Cost-benefit analysis of datacenter consolidation using virtualization." IT Professional, 2014.
Doriya, Rajesh, Pavan Chakraborty, and G. C. Nandi. "'Robot-cloud': a framework to assist
heterogeneous low cost robots ." International Conference on Communication, Information, and Computing Technologies. 2012. 1-5.
"Enabling a new future for cloud computing." National Science Foundation. August 20, 2014.
http://www.nsf.gov/news/news_summ.jsp?cntn_id=132377 (accessed 10 31, 2015).
Gungor, V.C., and G.P. Hancke. "Industrial wireless sensor networks: challenges, design principles, and technical approaches." Industrial Electronics, IEEE Transactions on 56, no. 10 (2009): 4258-4265.
Hempel, Ralph, and David Lechner. ev3dev. n.d. http://www.ev3dev.org/ (accessed October 2, 2015).
Kehoe, Ben, Sachin Patil, Pieter Abbeel, and Ken Goldberg. "A survey of research on cloud robotics and automation." IEEE Transactions on Automation Science and Engineering, April 2015: 398-409.
Kelmendi, Rilind. "Economics of cloud computing latency." International Journal of Academic Research 5, no. 2 (2013): 99-101.
LEGO Education EV3. 2015. https://education.lego.com/en-us/lesi/middle-school/mindstorms-education-ev3/all-about-ev3 (accessed 10 18, 2015).
Monajjemi, Mani. "ardrone_autonomy." GitHub. April 24, 2015.
Nguyen, Vu-Anh-Quang, and Myungsik Yoo. "Packet loss compensation for control systems over
industrial wireless sensor networks." International Journal of Distributed Sensor Networks, 2015:
1-9.
Parker, Lynne E., and Fang Tang. "Building multirobot coalitions through automated task solution synthesis." Proceedings of the IEEE. 2006. 1289-1305.
Parrot AR.Drone2.0. 2015. ardrone2.parrot.com (accessed 10 18, 2015).
Pye, Andy. "THE INTERNET OF THINGS connecting the unconnected." Engineering & Technology (17509637), December 2014: 64-70.
Quidley, Morgan, et al. "ROS: an open-source Robot Operating System." Willow Garage. 2009.
https://www.willowgarage.com/sites/default/files/icraoss09-ROS.pdf (accessed 10 25, 2015).
ROS/Concepts. Jun 21, 2014. http://wiki.ros.org/ROS/Concepts (accessed 10 18, 2015).
Sharp, Ron. "Latency in cloud-based interactive streaming content." Bell Labs Technical Journal (John Wiley & Sons, Inc.) 17, no. 2 (2012): 67-80.
"The NIST definition of cloud computing." NIST. September 2011.
http://csrc.nist.gov/publications/nistpubs/800-145/SP800-145.pdf (accessed November 18, 2015).
Turtlebot 2. n.d. http://www.turtlebot.com (accessed 10 18, 2015).
Wang, Lujia, Ming Liu, and Max Meng. "Real-time multisensor data retrieval for cloud robotic systems."
IEEE Transactions on Automation Science and Engineering 12, no. 2 (2015): 507-518.
Weir, Bill. "Resolving signal integrity issues in cloud computing platforms." ECN: Electronic Component News, May 15, 2013: 16-17.
Whitson, George M., III. "Computer Networks." Salem Press Encyclopedia of ScienceResearch Starters.
2015.
Willcocks, Leslie P., Will Venters, and Edgar A. Whitley. Moving to the cloud corporation: how to face the challenges and harness the potential of cloud computing. Palgrave Macmillan, 2013.
Appendix A LIST OF PUBLICATIONS
C. Reid and B. Samanta, “Gesture Recognition for Control in Human-Robot Interactions,” in ASME 2014 International Mechanical Engineering Congress & Exposition, Montreal, 2014.
C. Reid, B. Samanta, and C. Kadlec, “Development of a Network Infrastructure for Heterogeneous Robot and Control Systems Interactions,” in ASME 2014 International Mechanical Engineering Congress & Exposition, Houston, 2015.
C. Reid, B. Samanta, and C. Kadlec, “Heterogeneous Networked Systems in a ROS-Enabled Cloud Environment,” in 2016 ASNE Symposium Proceedings, Arlington, 2016.